One principle has governed science funding in the United States for decades: that scientific judgement should not be supplanted by political pressures. That is the basis upon which the country’s enormous scientific achievements—progress in lifesaving treatments, a university system that is the envy of the world, and a vibrant technology-and-biotech sector—have been built.But that principle is under attack. Late last month, the Office of Management and Budget in the Donald Trump White House proposed a new regulation that would shift control over the allocation of public science funding away from scientists and toward political appointees, who will have the power not only to decide which projects receive that money but also to, at any time, cancel grants that have already been awarded.This isn’t merely an attack on science; it is part of a deeper ideological vision that many within the Trump administration have been championing for years: moving power from Congress and executive-branch agencies to the president and his appointees. This vision of increased presidential power permeates Project 2025—a plan co-authored by Russell Vought, now the OMB head. And it is backed by a legal idea known as the unitary-executive theory, which asserts, as Justice Antonin Scalia put it in 1988, that the president possesses all federal executive power. The unitary-executive theory, in its fullest form, means that the president can review and direct all executive-branch actions and that Congress cannot restrain that power unless otherwise laid out in the Constitution.[Alexander Furnas and Dashun Wang: The Republicans made peace with science]The Supreme Court has in recent years issued a series of decisions in line with the unitary-executive theory, many of which are about the president’s power to fire agency officials. The same theory is also behind numerous Trump-administration actions taken in apparent conflict with congressional action or intent, such as the president’s orders to dismantle the Department of Education and bulldoze the East Wing of the White House to build a ballroom. The administration is fighting a battle over what Congress in legislation can direct agency experts to do, versus what the president has the power to control. Now the proposed OMB rule would extend unitary-executive logic to grantmaking, including at America’s science agencies, the foundation of its medical and industrial achievements.Elias Zerhouni, the National Institutes of Health director under President George W. Bush, earlier this year described what’s at stake. “When I was at NIH, it never occurred to me that I could say we’re not paying for something that Congress had appropriated money for because it was not in line with my policies,” he said. “But if that power is given to the executive branch to use whenever it wants, then all bets are off.”I have worked at NIH for more than a decade, running a neuroscience lab that investigates how the trillions of connections in brains control thought and action, work aimed at understanding diseases such as schizophrenia and autism. Before last January, scientists like me could set our own scientific goals and follow data where they led, subject to regular evaluation by a panel of external scientific leaders, who would assess our progress and the value of our work. Now what I see in all of the changes reported at NIH this year—speaking for myself, not the agency—is the addition of presidential control and review steps at many levels and decision points. The proposed OMB rule in fact would formalize many things already happening inside NIH, such as grant cancellations, political review of grant awards, control over staff on panels that review scientific programs, even the forced alteration of words in grant proposals. An extreme assertion of executive power is bad enough when the issue is a ballroom; it is far more destructive when the target is American science.Advocates for presidential control of agencies—including Chief Justice John Roberts—have argued that the president must have fine-grained control over the whole executive branch, because only the president (along with the vice president) has the accountability of being elected by the whole nation. But the science agencies’ great success has come while they have drawn democratic accountability from the public via Congress. The legislative process has set long-term agency priorities, and civil servants at the agencies have carried them out. The president has played a constitutional role in signing or vetoing legislation affecting the agencies, but beyond that, his power has been indirect: as an advocate who could pressure Congress but who did not take action directly himself.For example, in 1971, when President Richard Nixon wished to launch a “war on cancer,” he did not do so by directing NIH to change its priorities. Instead, he announced his plan at a State of the Union address, and then worked with Congress to enact a law, the National Cancer Act, that established a national research effort. The president did have the power to appoint the head of the National Cancer Institute (as well as the head of the NIH), but he did not exert control over the minutiae of the agencies’ work.Beyond their accountability to the public via the legislature, science agencies such as NIH and the National Science Foundation have also been accountable to another constituency: the scientific community. NIH and NSF rely heavily on advisory input from outside scientists. Peer-review panels with rotating expert members perform review of grants. (In 2023, nearly 30,000 people served as NIH peer reviewers.) External scientists invest large amounts of their time in careful reviews, and those reviews heavily shape funding decisions. Advisory councils and ad hoc committees of external scientists approve new research priorities, provide a second-level review of grants, and advise the agency in setting scientific directions.Drawing on the expertise of the scientific community for funding decisions has been a key part of U.S. scientific success. Just as in economics, in which markets allow the decentralized knowledge of many individuals to be harnessed, in science, relying on peer review allows the expertise of many scientists to be brought to bear. Central planning of science via the White House and political appointees would cast aside the enormous collective expertise of the American scientific community.As the Trump administration has reorganized U.S. science, it has justified the need for change by citing long-standing critiques. For example, critics of grant peer review, even before this administration, have noted that the practice can lead scientists to be conservative and avoid pursuing high-risk, blue-sky work—because getting a group of reviewers to agree on the value of a risky idea is hard. But I know from my own experience writing and reviewing grants—and a team including the former NIH director Harold Varmus noted as much more than a decade ago—that the factor that cuts most acutely against big ideas is low funding rates. When 30 percent of grant applications were funded, there was plenty of room for peer reviewers to have differing views, and for review panels to endorse some moonshot projects alongside safe projects. Now, because the success rate has fallen in some cases to 5 percent, near unanimity among reviewers is required, and only the projects that reviewers see as most certain of success will receive a green light. Shifting the science system to presidential control will do little to address this issue. What’s more, big discoveries are less likely to come from projects chosen by political appointees, who are prone to picking flashy and well-marketed work instead of unusual-sounding but promising ideas that experienced scientists could identify.The OMB’s proposed rule has already created outcry in the science world. Scientific societies have condemned the rule and called on their members to comment. Holden Thorp, the editor in chief of Science, said in an editorial that “the time to act is now. The scientific community needs to flood OMB with responses.”[Hana Kiros: The Trump administration is done with social science]But returning control of scientific decision making to the scientific community will not be simple. The proposed rule is just one step in the Trump administration’s efforts to put science under the president’s thumb, and is aligned with the Roberts Court’s vision. The unitary-executive theory implies that presidential appointees must have final say over agency decisions, no matter what laws Congress passes.America is in the middle of a major shift in what the separation of powers means. A president focused on taking power for himself is backed by a Supreme Court majority that has supported that agenda. And this means that restoring American science’s vitality will need bigger thinking than just a response to this proposed rule. Scientists and allies must create a new vision for how science agencies should be run, and that vision must itself be part of a bigger rethinking of how the American government should be run. The Roberts Court is unlikely to agree with that approach, but it is worth fighting for—not just for science’s sake, but for America’s too.